A natural experiment of dietary overlap between introducedRainbow Trout (Oncorhynchus mykiss) and native Puyen(Galaxias maculatus) in the Santa Cruz River, Patagonia

Marina Tagliaferro & Ivan Arismendi &Julio Lancelotti & Miguel Pascual

Received: 3 February 2014 /Accepted: 16 October 2014 /Published online: 22 October 2014# Springer Science+Business Media Dordrecht 2014

Abstract Diet overlap between the native Puyen(Galaxias maculatus) and juvenile exotic Rainbow Trout(Oncorhynchus mykiss) was studied in 52 sites locatedalong 306 km of the mainstem of the Santa Cruz River,one of the largest rivers in Patagonia. The relative abun-dance of both species varied along the river, with threeclearly defined areas including an upstream “highRainbow Trout to Puyen ratio” area (with abundances of75 and 25 %, respectively), a midstream “intermediateRainbow Trout to Puyen ratio” area (relative abundancesbetween 75 and 25 %), and a downstream “low RainbowTrout to Puyen ratio” area. The diet of the 2 species wasanalyzed across these 3 areas examining stomach content.Diet similarity between species was analyzed using a non-metricmultidimensional scaling ordination technique; preyelectivity was evaluated with the Ivlev’s Index; feedingtactics were studied by estimating prey-specific abun-dance. Both species showed a generalist feeding tactic,with Puyen exhibiting a more varied diet. Prey electivitywas similar in both species, with the mayfly (Meridialaris

chiloeensis), stoneflies (Klapopteryx kuscheli andAntarctoperla michaelseni), and the amphipod (Hyalellasp.) being the most frequently consumed prey. A signifi-cant diet overlap was found only in the downstream areaswhere a higher proportion of native fish occurs. The lowdiet overlap in upstream locations might be because of thehigh density of Rainbow Trout; while mid-stream could bedue to the high secondary productivity spots. Our resultssuggest that the diet of native Puyen changed in relation tothe abundances of Rainbow Trout in the stream.

Keywords Trout invasion .Galaxiids .n-MDS .Feedingtactics

Introduction

Major decreases in the biodiversity of aquatic systemsduring the last century have been attributed to habitatloss and species introduction (Mack et al. 2000; Muotkaand Syrjänen 2007). Salmonids (genus Oncorhynchus,Salmo and Salvelinus) are among the most widely in-troduced fish taxa around the globe (Welcomme 1988),mostly for recreational and aquaculture purposes(McDowall 1994; Pascual and Ciancio 2007).Salmonid introductions have been broadly implicatedin the decline of native biota (Crowl et al. 1992; Greigand McIntosh 2006; Soto et al. 2007; Arismendi et al.2009), with consequences at all levels of ecologicalorganization including behavior alterations at indi-vidual level (Simon and Townsend 2003), reduc-tion in population abundance (Moyle and Light

Environ Biol Fish (2015) 98:1311–1325DOI 10.1007/s10641-014-0360-6

M. Tagliaferro (*)División de Ciencias Básicas - Universidad Nacional deLuján. Ruta 5 y Av. Constitución,Luján (6700) Buenos Aires, Argentinae-mail: azulmarinita@gmail.com

I. ArismendiDepartment of Fisheries and Wildlife, Oregon StateUniversity,3200 SW Jefferson Way, Corvallis, OR 97331, USA

J. Lancelotti :M. PascualCONICET, Centro Nacional Patagónico,Bld. Brown 2915, Puerto Madryn, Chubut, Argentina

1996; Townsend 2002), and trophic cascades(Simon and Townsend 2003).

Starting in the early 20th century, and pushed by thehigh value of salmonids for fisheries and aquaculture,governments in Chile (Basulto 2003), New Zealand(McDowall 1990; Flecker and Townsend 1994), andArgentina (Tulian 1908; MacCrimmon 1971; Lever1996) promoted the introduction and establishment ofsalmonids in the Southern Hemisphere. More than tensalmonids species were introduced in southern rivers inChile (Arismendi et al. 2014) and Argentina (Pascual et al.2002). However, it has only been during the last 30 yearsthat the ecological consequences of salmonid establish-ment in these areas started to be observed and investigatedwith a special focus on their effect on native communities(e.g., Crowl et al. 1992; Arismendi et al. 2009; Young et al.2010), the resources used (Kusabs and Swales 1991; Buriaet al. 2007; Penaluna et al. 2009), competition and preda-tion (McIntosh et al. 1992; Macchi et al. 1999; McDowall2003), and niche overlap (Vargas et al. 2010; Arismendiet al. 2012; Correa et al. 2012; McHugh et al. 2012).

The study of diet breadth between the exotic andselected native species provides a practical entry pointto the understanding of the impacts of salmonids inPatagonia (Soto et al. 2007; Habit et al. 2010).Whereas niches may be characterized as measures ofresource utilization (Giller 1984), diet breadth refers tothe utilization of some of the same type of resources bytwo ormore species of resource consumers (Colwell andFutuyma 1971; Abrams 1980). In particular, exotic spe-cies provide natural experiments where the ecologicaltheory (e.g. optimal foraging after an invasion) may betested empirically through evaluating habitat segrega-tion, density reduction, niche shifts, food intake reduc-tion, prey composition alteration, or size structurechanges (Bøhn et al. 2008), with possible temporalshifts in feeding habits (Coghlan et al. 2007).

Experimental manipulations of the density of exoticspecies arguably provide the most powerful approach tostudy system-level effects of single species on native com-munities (Tilman 1987; Hansson et al. 1998). However,rivers are large and open systems where fish are highlymobile and thus, manipulation is often very difficult.Researchers are then limited to looking for natural exper-iments, where naturally contrasting densities occur. Thisapproach has been applied, for instance, to study the effectsof Rainbow Trout (Oncorhynchus mykiss) on galaxiids(Galaxias spp.) niche width (Townsend 2002; McHughet al. 2012) and the interaction for food and space between

populations of galaxiids and juvenile trout in NewZealand(Glova et al. 1992; McIntosh 2000; McDowall 2003;McIntosh et al. 2010). It has also been used in SouthAmerica to study habitat use and segregation of nativefishes and trout in Chilean streams (Penaluna et al. 2009)and differential piscivory effects on galaxiids by nativepredators and trout in Northern Patagonia (Arismendiet al. 2012; Juncos et al. 2013).

One of the main deficiencies in our current knowledgeconcerning exotic trout in southern Patagonia is abouttrophic relationships and interactions with conspicuousnative fishes such as Galaxiids and Siluriforms (Pascualet al. 2002). The aim of the present study is to evaluatediet overlap of underyearling Rainbow Trout(Oncorhynchus mykissWalbaum 1792) in sympatry withthe most abundant native species Puyen (Galaxiasmaculatus Jenyns, 1842), in the Santa Cruz River, oneof the largest rivers in the region. We concentrated oureffort during the low flow season (springtime) along306 km of themainstem of the Santa Cruz River, betweenthe head lake, Lago Argentino, and the outflow into theAtlantic Ocean (Fig. 1). We evaluated if Rainbow Troutand Puyen are using similar food resources along agradient of contrasting relative abundance of both speciesas a preliminary exploration of mechanisms of competi-tion avoidance.

Since no information regarding native galaxiids distri-bution and feeding habits before trout invasion is availablefor large rivers, this study represents an important firstapproximation to examine interspecific interactions be-tween salmonids and native fishes in this understudiedregion. Because the Santa Cruz River can be considereda minimally human-influenced river (Brunet et al. 2005;Tagliaferro et al. 2013), our study will also provide abaseline for future comparison with other large rivers.

Materials and methods

Study area

The Santa Cruz River (50° S; 70° W) originates in twooligotrophic to ultra-oligotrophic large glacial lakes,Viedma and Argentino, and flows uninterrupted for382 km across the Patagonian plateau to drain into theAtlantic Ocean (Fig. 1; Brunet et al. 2005). The river hasan average discharge of 691m3 s−1 (min. 278.1 m3 s−1 inSeptember and max. 1,278 m3 s−1 in March), which ishighly predictable due to a glacial dominated regime

1312 Environ Biol Fish (2015) 98:1311–1325

(Tagliaferro et al. 2013). This river has low variability ofthe environmental structure (Tagliaferro et al. 2013); it isan un-braided river (100–200 m wide×382 km length)and temperature between upstream/ downstream areasdiffers only by 3–5 °C at a given time of the year. Alongthe river, temperature varies only in 2 to 5 °C during aperiod of few weeks, e.g. temperature variation duringsampling was 5.7±0.5 °C, 7.0±0.9 °C, and 8.6±0.9 °Cfor upstream, midstream and downstream areas respec-tively). The Santa Cruz River has been characterized asthe one of the poorest in terms of macroinvertebrateabundance among 40 Patagonian rivers (Miserendino2001).

Sampling

We sampled fish and macroinvertebrates duringSeptember 2010 (low flow period) in 52 sites at intervalsof 6 km along the Santa Cruz River (50° S; 70° W,

Fig. 1). The uppermost site was located in Charles Fuhr(9.8 km downstream from the LakeArgentino, 50°16′ S;71°53′ W) and the lowermost site was located inComandante Luis Piedra Buena (50° S; 70°60′W), closeto the river’s estuary and 315.8 km from LakeArgentino. We captured fish (length range: 50–140 mm) using standard single-pass electrofishing pro-cedures from littoral zone to depths of 0.6 m (Jones andStockwell 1995; Meador et al. 2003). The equipmentused was a Smith-Root LR-24 electrofisher set to afrequency of 90 Hz and a pulse width of 3 ms. At eachsite, a coastal wadable stretch of 100 m was sampledfollowing a zig-zag track. Catchability may have dif-fered between fish species, but we assumed it to besimilar throughout sample sites. All fish were counted,fork length-measured with a digital caliper (0.01 mmnearest unit), and weighed on a Mettler PC 440 DeltaRange balance (0.003 g nearest unit). We used thenumber of individuals captured in the 100 of river as

Fig. 1 Map of the Santa Cruz River, Argentina. Sampling sites are located between Charles Fuhr Bridge and Comandante. Luis PiedraBuena Town (between arrows)

Environ Biol Fish (2015) 98:1311–1325 1313

an indirect measurement of abundance (CPUE).Stomachs from 5 to 10 randomly selected fish from eachspecies and si te were removed and stored.Macroinvertebrate samples were obtained with a kick-net of 450 μm mesh size covering 0.25 m2, integratingone area for each sample (Tagliaferro et al. 2013). Driftsamples were not included in the analyses because thebiomass contribution to the total macroinvertebrate wasless than 2 % along the river (Tagliaferro 2014).Macroinvertebrate samples were stored in a portablefreezer at −18 °C. At the laboratory, both stomachcontents and macroinvertebrate samples were trans-ferred into 70 % ethanol for further separation andidentification of organisms to the lowest possible taxo-nomic level, employing a Zeiss stereomicroscope(6.5 X). We identified taxa following Lopretto and Tell(1995), and Domínguez and Fernández (2009).

We measured free-living macroinvertebrates richnessand abundance, and dry weight of consumed prey forthe data analysis. To calculate dried weight, we assignedindividuals to a given taxon, dried to a constant weightat 65 °C (24 h), and weighed on an analytical ShimadzuAUW-220 scale (range: 220 g – 10 mg, error: 1 mg).When prey was partially digested, weight was estimatedfrom complete items of the same length. Chironomidaespecies, which were similar in size and ecological role,were pooled as one group. Other small taxa (e.g.Oligochaeta, Glossosomatidae, and Trichoptera) werealso pooled as one group.

Fish relative abundances were estimated for eachsampling site to assess patterns of distribution ofRainbow Trout and Puyen. Three different sections ofthe river were defined based on the Rainbow Trout toPuyen ratio: an “upstream” area, with high proportion ofRainbow Trout (over 0.75), a “downstream” with lowproportion (less that 0.25), and a “mid-stream” withintermediate proportion (0.25 and 0.75). Stomach con-tents of 17 Puyen and 32 Rainbow Trout were analyzedin the upstream area; while in mid-stream, 82 Puyen and116 Rainbow Trout, and in downstream area 101stomachs of Puyen and 44 of Rainbow Trout wereexamined.

Fish diets

We used three methods to characterize the fish diets(Hyslop 1980; Chipps and Garvey 2007): including (a)the frequency of occurrence (%Fi) of a given prey type,defined as the mean number of stomachs in which that

prey occurs in relation to all the stomach studied foreach area; (b) the biomass contribution (%Bi) of a preyto the dry weight of each stomach contents, and (c) preyspecific abundance (Pi; Amundsen et al. 1996). Theseindices were calculated as follow:

%Fi ¼ Ni

N

%Bi ¼X

i

t SiSt

!*100

Pi ¼X

i

t SiSti

!*100

where Ni is the number of fish with prey i in theirstomach, N is the total number of fish with stomachcontents, Si is the contribution of prey i to stomachfullness in dried mass, St is the total stomach biomassof all fish, and Sti is the total stomach fullness of fishwith prey i in their stomach. At each area (upstream,mid-stream, and downstream), Fi and Pi were used tocreate a diagram of “feeding tactics” (Amundsen et al.1996).

Multivariate analysis of diets

We used a non-metric multidimensional scaling (N-MDS) ordination technique to compare the similarity ofdiets between species and across areas using the Bray–Curtis distance metric (Clarke 1993; Marshall and Elliott1997). Based on an iterative optimization procedure, dietcompositions were rearranged to minimize a measure ofdisagreement or stress between their distances in 2-D(Kruskal 1964). The resulting coordinates of each pointfrom the 2-D plot provided a collective index of howunique the diet of a given fish was. The proximity ofpoints in a 2-D plot indicates a higher degree of similarity,whereas more dissimilar points are positioned furtherapart. We used the standard squared root transformationto down-weight the importance of the highly abundantpreys (see Clarke and Warwick 2001 for more details).We used the software PRIMER v6.1.5 (Clarke andGorley 2006) to produce the ordination plot of theBray–Curtis similarity coefficient of square root trans-formed % frequency of prey for each individual(Rainbow Trout or Puyen) and % biomass of each preyat each fish species (Clarke and Warwick 2001).

We tested the hypothesis of no difference amonggroups (fish species and areas) of diets using an analysisof similarity (ANOSIM). The ANOSIM is a

1314 Environ Biol Fish (2015) 98:1311–1325

Tab

le1

Macroinvertebrate

naturalabundance(%

D),frequencyof

occurrence

(%Fi)andprey-specificabundancein

biom

ass(Pi-B)of

preysin

stom

achcontentsof

O.mykissand

G.m

aculatus;v

aluesarecalculated

ateach

area

Upstream

Mid-stream

Dow

nstream

Trout

Puyen

Trout

Puyen

Trout

Puyen

Taxa

%D

%Fi

%Bi

Pi%Fi

%Bi

Pi%D

%Fi

%Bi

Pi%Fi

%Bi

Pi%D

%Fi

%Bi

Pi%Fi

%Bi

Pi

Mollusca

Chilin

asp.(Ch)

22.5

––

––

–6.5

0.3

19.9

3.7

0.2

13.6

0.7

––

––

––

Heleobiasp.(He)

––

––

––

––

––

––

5.1

2.4

0.9

37.8

––

–

Lymnaea

sp.(Ly)

–2

0.6

213

3.6

23.9

389.7

4.8

35.7

14.8

0.8

41.8

52.9

2.3

2.6

79.5

4.0

4.9

30.4

Annelida

Hirudinea

(Hir)

0.1

––

––

––

0.3

––

–3.7

0.3

18.7

1.8

––

––

––

Naididae(N

)0.2

––

–2

0.1

2.06

0.1

3.2

0.9

100

7.4

0.8

3.6

––

––

4.0

0.3

2.0

Acari

Acarispp.(A

c)0.1

––

–<0.1

0.1

1.54

––

––

––

––

––

–4.0

0.2

1.0

Crustacea

Hyalella

spp.(H

)22.6

4213.4

21.0

45.0

31.4

20.5

1871.0

21.2

37.4

55.6

24.1

33.5

17.4

35.0

9.5

21.5

20.0

5.7

17.0

Ephem

eroptera

Andesiops

sp.(Ad)

0.2

195.1

18.0

16.0

5.5

17–

9.7

2.8

4.0

7.4

1.5

3.7

0.3

10.0

1.5

11.0

4.0

1.3

16.5

Meridialarischilo

eensis(M

c)19.9

6329

38.5

17.0

10.4

9811

9.7

13.1

25.6

7.4

15.0

19.2

4.4

57.0

21.6

38.5

24.0

9.1

41.0

Plecoptera

Aubertoperlailliesi(A

i)5.1

50.6

22.0

––

–0.6

54.8

1.9

74.4

––

–1.7

––

–16.0

<0.1

13.5

Antarctoperla

michaelseni

(Am)

0.0

2412

42.0

23.2

10.0

22.1

–25.8

8.5

32.2

33.3

5.9

35.2

0.1

46.1

21.4

42.0

24.0

14.5

86.3

Klapopteryx

kuscheli(K

k)0.2

2213

84.0

16.0

8.8

79.4

4.9

6.5

16.4

65.3

11.1

10.6

76.9

0.2

16.2

8.9

67.3

8.0

8.1

92.0

Limnoperlajaffu

eli(Lj)

–38

9.1

7.3

15.0

8.8

332.5

61.3

14.1

12.3

37.0

10.7

14.3

6.1

84.3

21.5

16.3

63.0

29.1

38.9

Coleoptera

Luchoelmiscekalovici(Lc)

14.2

122

4.5

11.0

2.5

2819

48.4

6.7

7.9

25.9

7.3

4.3

8.3

18.0

3.0

1.3

8.0

0.9

11.3

Luchoelmiscekaloviciadult(elA)

––

––

3<0.1

8.84

–3.2

–2.7

––

––

––

––

––

Veliidae

––

––

––

––

––

–14.8

0.3

72.7

––

––

––

–

Trichoptera

Mastig

optilasp.(M)

0.3

12–

7.2

2.5

0.2

2.06

–19.4

0.3

2.0

3.7

0.2

0.9

––

––

––

–

Rheochorema/

Atopsyche

sp.(Hy)

0.2

203.5

16.7

18.0

2.1

17–

22.6

1.9

8.4

11.1

1.8

12.6

0.3

26.0

6.7

17.0

8.0

1.7

2.8

Caillo

masp.(C)

0.0

42

0.4

7.0

3.7

6–

–0.4

––

––

–3

0.2

8–

––

Smicrideadithyra(Sd)

2.2

80.1

2.1

8.0

0.8

7.66

1.2

6.5

0.1

4.5

––

––

–0.1

––

––

Environ Biol Fish (2015) 98:1311–1325 1315

Tab

le1

(contin

ued)

Upstream

Mid-stream

Dow

nstream

Trout

Puyen

Trout

Puyen

Trout

Puyen

Taxa

%D

%Fi

%Bi

Pi%Fi

%Bi

Pi%D

%Fi

%Bi

Pi%Fi

%Bi

Pi%D

%Fi

%Bi

Pi%Fi

%Bi

Pi

Oxyethira

sp.(O)

0.0

––

–<0.1

<0.1

3.15

––

––

3.7

0.1

9.4

––

––

––

–

Dyctisidae

––

––

––

––

––

–3.7

0.6

18.3

––

––

––

–

Diptera

Chironomidae

11.6

00.1

2.6

9.4

5.8

214

–0.1

–25.9

1.4

8.7

0.2

0.3

<0.1

1.4

23.1

8.9

17.9

Eukiefferiella

sp.(Eu)

Paratrichocladius

sp.(Pcl)

Endotribelossp.(En)

Parachironomus

sp.(Pch)

Alotanipussp.(Al)

Other

larvae

0.1

––

––

––

0.2

––

––

––

––

––

4.0

0.3

2.3

Pelecorhynchidae

(Pe)

Empididaesp.(Em)

Muscidaesp.(Mu)

Simuliid

ae0.6

312.4

1.9

22.5

6.2

110

25.8

1.1

1.8

11.1

1.1

6.0

0.4

36.0

2.1

4.3

28.0

7.0

9.7

TerrestrialP

reys

Trichoptera

0.0

167

46.2

––

––

–3.2

–33.3

13.0

32.2

–4

<0.1

594.0

7.4

79.0

Ephem

eroptera

––

––

––

––

––

–7.4

3.9

40.9

––

–<0.1

0.8

2.5

Formicidae

––

––

––

––

––

––

––

––

––

4.0

0.2

3.8

1316 Environ Biol Fish (2015) 98:1311–1325

nonparametric procedure analogous to analysis of vari-ance (ANOVA) that tests differences in distance in theordination matrix of fish diets against random groups (adetailed procedure is provided by Clarke 1993). Weconducted 99,999 random permutations to estimate thesignificance of the R test statistic associated withANOSIM and the posteriori pairwise tests. The Rstatistic ranges between −1 and +1, where −1indicates more similarity between groups thanwithin groups, 0 indicates no effect of groupsand +1 indicates less similarity between groupsthan within groups. Though an R value of 0.45is expected when groups exhibit significant differ-ences; when analyzing a large number of repli-cates, R value could be smaller and statisticallysignificant (Clarke 1993). In addition, we per-formed a similar procedure to evaluate possibledifferences in macroinvertebrates abundances andbiomass among upstream, mid and downstreamareas.

A SIMPER, or ‘similarity percentage’, analysis wasperformed on squared-root transformed data providing aranking that shows which prey items contributed mostby percentage to the similarity in a within group test, orthe percentage of dissimilarity contributed to a betweengroup test. SIMPER decomposes average Bray– Curtissimilarities between all pairs of samples in groups (orbetween groups of samples) into percentage contribu-tions from each dietary item (Clarke 1993). Finally, thesimilarity within groups was ranked by estimating theindex of multivariate dispersion, which provides infor-mation on diet width (MVDISP; Warwick and Clarke1993).

Degree of dietary overlap and prey electivity

As a complement of the N-MDS andANOSIM analysis,we calculated the degree of the dietary overlap betweenPuyen and Rainbow Trout at each area using theSchöener index (1970):

CPuyen=trout ¼ 1−0:5*X

pPuyen;i−ptrout;i�� ��� �

;

where p,puyen,i, is the proportion of prey i in Puyen,and ptrout,i, is the proportion of prey i in Rainbow Trout.FollowingWallace (1981), we considered aCPuyen/trout=0.6 or above as a significant diet overlap. A value of 0.6indicates a 60 % of overlap in the diet.

In addition, we calculated the Ivlev electivity index(Ivlev 1961) as a measurement of fish electivity of preyas follows:

I i ¼ ri−piri þ pi

where ri and pi are the proportion of item or prey i inthe stomach content and environment, respectively. Therange of Ii varies from −1 (complete rejection) to +1(positive chosen), with values around zero indicatingthat feeding is proportional to item abundance in the

Fig. 2 Non-metric MDS ordination plot (2D) showing the degreeof similarity of prey density (upper panel; stress 0.17) and preybiomass (lower panel; stress 0.17) among the 52 sites grouped byup- mid and downstream areas using Bray-Curtis similarity index.Proximity of symbols indicates a higher degree of similarity

Environ Biol Fish (2015) 98:1311–1325 1317

environment. We considered positive or negativeelectivity when Ii was above 0.5 and below −0.5respectively.

Results

A total of 587 Rainbow Trout and 1,183 Puyen werecaptured with fork length 40–90 mm and weight 0.5–9.0 g, similar for both species along the 306 km of river.A total of 192 and 200 stomach contents of RainbowTrout and Puyen, were analyzed respectively. Twenty-eight prey types for Rainbow Trout and 38 for Puyenwere found, from autochthonous and allochthonous(terrestrial) origin. A total of 38 exclusive benthic mac-roinvertebrates were found within all the sampling sites(Table 1).

ANOSIM showed a significant dissimilarity amongareas (for prey density R statistic=0.059, P=0.001; forprey biomass R statistic=0.071, P=0.009). PairwiseANOSIM results for prey density showed significant sta-tistical differences between upstream andmid-stream areas(R statistic=0.11, P=0.001), and between mid-stream anddownstream areas (R statistic=0.09, P=0.006), but not forupstream versus downstream areas (R statistic=0.001, P=0.34). Moreover, pairwise ANOSIM results for prey bio-mass showed significant statistical differences betweenupstream and mid-stream areas (R statistic=0.12,P=0.006), and between mid-stream and down-stream areas (R statistic=0.08, P=0.008), but not for

upstream versus downstream areas (R statistic=0.05,P=0.12). Mid-stream areas exhibited a greater variabil-ity with many sites with high macroinvertebrates bio-mass, a possible reason why the non-metric MDS ordi-nation plot showed the distribution overlap of preydensity and prey biomass among the 52 sites groupedby up- mid and downstream areas using Bray-Curtissimilarity index (Fig. 2).

Because the distribution of fish showed high predom-inance of Rainbow Trout from upstream towards mid-stream sites, and decreased in downstream sites, wherePuyen were more predominant, (Fig. 3) diet overlap wascompared among areas. In upstream areas,Rainbow Trout-Puyen ratio ranged from 0.75 to1, where Puyen exhibited the lowest CPUE valuesfrom 0 to 13 ind. per 100 linear meters. In down-stream areas, Rainbow Trout-Puyen ratio rangedfrom 0.1 to 0.25. The mid-stream area showed aRainbow Trout-Puyen ratio ranging between 0.25to 0.75 with CPUE values of 22±17 ind. per 100linear meters and 25±17 ind. per 100 linear meters forRainbow Trout and Puyen respectively. We assumedthat the species-specific catchability remained un-changed along the river.

Diet overlap

A higher degree of dietary similarity between RainbowTrout and Puyen was found in both biomass and numberof prey only for the area downstream, (diet overlap for

Fig. 3 Capture per unit of effort(CPUE; fish per 100m of river) ofPuyen (G. maculatus) andRainbow Trout (O. mykiss) at the52 sampling sites located 6 kmfrom each other and grouped bythe three up- mid and downstreamareas. Box-plots represent 5th,25th, median, 75th and 95th per-centiles. Numbers above eachbox-plot indicates the range offish ratios at each area

1318 Environ Biol Fish (2015) 98:1311–1325

%Fi and %Bi; P=0.19; Fig. 4; Table 2). The other twoareas, upstream and mid-stream, showed statisticallysignificant differences in the similarity of diets betweenRainbow Trout and Puyen. Also, each species’ dietvaried across river areas, both in biomass and numberof prey (Table 2, bottom part). Of all species-specificpairwise comparisons between areas, only those forRainbow Trout between upstream and downstream

areas were non-significant. Even small visual differ-ences appeared in the N-MDS plot (Fig. 4), there wasa statistically significant difference in the diet betweenspecies in sites both upstream and midstream (for %Fiand %Bi R statistic ranged between 0.09 and 0.18,P<0.001; Table 2).

The degree of overlap in diets between Puyen andRainbow Trout, estimated as the Schöener index,showed a differential response among areas, whichwas concordant with the N-MDS and ANOSIM analy-ses. Lowest values were found upstream at 20 to 30 %overlap detected, while the range of values in the mid-stream area was intermediate (25 to 45 %), indicating alow diet overlap between species at that area. Thehighest Schöener index was found downstream wherethe Rainbow Trout presented the lowest abundance,including sites with 20 to 80 % diet overlap.Despite the large amplitude range, only this areaexhibited sites with diet overlap of up to 80 %between the two species. Therefore, high dietoverlap occurred between Puyen and RainbowTrout only in the downstream area with overlapdegrees of up to 80 %.

The SIMPER analysis showed that the amphipodHyalella sp., the stonefly Limnoperla jaffuelli, the may-fly Meridialaris chiloeensis, and Simuliidae larvaeaccounted for most of the dissimilarity in diet betweenspecies for both upstream and mid-stream areas(Table 3). While in mid-stream areas the elmidLuchoelmis cekalovici was also a prey that significantlycontributed to diet differences between Rainbow Troutand Puyen (13.5 %), in upstream areas, the stoneflyAntarctoperla michaelseni accounted for 8.2 % of thedifferences. Generally, a higher number of prey contrib-uted to the Puyen diet, which is consistent with a highermultivariate dispersion index (MVDISP). This indexindicates that the Puyen’s diet was more dispersed thanRainbow Trout in all three areas (for %Fi and %Bi).

Feeding tactics and prey electivity

Both Rainbow Trout and Puyen showed a generalistfeeding tactic in all three areas (Fig. 5). The frequencyof occurrence and prey-specific abundance of stomachcontents in Rainbow Trout and Puyen generallydisplayed a comparable pattern with both species con-suming similar proportions on different prey items.Diets of both Rainbow Trout and Puyen was character-ized by the occurrence of many rare prey and few

Fig. 4 Non-metric MDS ordination plot (2D) showing the degreeof similarity among individual fish diets of Puyen (G. maculatus)and Rainbow Trout (O. mykiss) from the up- mid and downstreamareas using Bray-Curtis similarity index. In the upper panel, indi-vidual fish diets are grouped by%Fi (stress=0.20) and in the lowerpanel by %Bi (stress=0.19). Proximity of symbols indicates ahigher degree of similarity

Environ Biol Fish (2015) 98:1311–1325 1319

consumed species frequently found such as the amphi-pod Hyalella sp., the mayfly Meriadialaris chiloeensis,the stoneflies Limnoperla jaffueli, Antarctoperlamichaelseni, the elmid Luchoelmis cekalovici, and adultmidges. These five taxa, with the exception of themidges, were the most widely distributed taxa inthe benthos along the river (Table 1) underlyingthe generalist feeding tactic by both species.Diptera, Simuliidae, and Chironomidae were thefamilies most frequently consumed by both fishspecies.M. chiloeensis, being one of the most recurrentpreys in Rainbow Trout across the three areas(Fig. 5) and resulted in a departure from the homogene-ity in the diet of both species.

Electivity of prey exhibited a similar general patternin both Rainbow Trout and Puyen (Fig. 6). Both speciespositively selected Simuliidae larvae, the stonefliesLimnoperla jaffuelli and Antarctoperla michaelseni, lar-vae of the caddisfly Hydrobiosidae and adult caddisfly.On the other hand, the elmid Luchoelmis cekalovici, thegastropod Lymnaea sp. and the caddisfly Smicrideadythira were avoided by both Rainbow Trout andPuyen. One of the most abundant macroinvertebrates,the amphipods, Hyalella araucana and H. curvispina,and the mayflyMeridialaris chiloeensiswere consumedin comparable proportion to environmental availabilityalong the river. In addition, Rainbow Trout positivelyelected the stonefly Klapopteryx kuscheli and the may-fly Andesiops sp., and avoided chironomids larvae;while the Puyen consumed these prey in similar propor-tions to those available in the stream.

Discussion

In agreement with previous studies conducted inPatagonia (Lattuca et al. 2008; Di Prinzio andCasaux 2012), we show that Rainbow Trout andPuyen are generalist predators, with a relatively lowelectivity of prey. The Puyen, however, appear tohave a wider diet breath than Rainbow Trout in theSanta Cruz River. The use of wider feeding sourcesin Puyen could explain the low diet overlap withRainbow Trout. This wider diet breath may be aresponse of native Puyen to feed on less desirableresources and thus, a response to a higher dominanceof Rainbow Trout. Differences in the diet of galaxiidswith and without Rainbow Trout were attributed tothe interaction with the latter (Glova et al. 1992;Glova and Sagar 1993; Elgueta et al. 2013).Rainbow Trout are known to exhibit a more vora-cious feeding behavior than galaxiids, which is con-sistent with the diet differences found in the areaswhere Rainbow Trout were more abundant.Alternatively to the competitive interactions basedon diet overlap includes predation on both species,the availability of suitable environment, and the riverproductivity to support the two species (Arismendiet al. 2012, 2014). However, there is no predationbetween the two species at this life stage. Indirectinteractions might occur such as habitat segregation(e.g. Penaluna et al. 2009) and it should be tested infuture comparisons. However, in the Santa Cruz Riverthere is not much cover (Tagliaferro et al. 2013;

Table 2 Pairwise ANOSIM results by species and areas

Pairwise Tests Biomass Number

Groups R- Statistic p level R- Statistic p level Significance

Between species

Upstream 0.17 0.003 ** 0.18 0.001 **

Mid-stream 0.12 < 0.001 *** 0.09 <0.001 ***

Downstream 0.03 0.190 0.02 0.265

Between areas

Trout Up vs mid-stream 0.13 0.007 ** 0.14 0.003 **

Up vs downstream 0.04 0.077 0.03 0.155

Mid vs downstream 0.08 0.024 * 0.14 < 0.001 ***

Puyen Up vs mid-stream 0.18 0.002 ** 0.18 0.002 **

Up vs downstream 0.20 0.003 ** 0.20 0.002 **

Mid vs downstream 0.16 < 0.001 *** 0.16 < 0.001 ***

1320 Environ Biol Fish (2015) 98:1311–1325

Quiroga pers. comm.) and therefore, we consider thatother possible interactions besides competition forfood might be less important.

Moreover, throughout the study area, the electivity ofprey is similar for both species, except in few prey suchas K. kuscheli, (one of the largest food items), or smallchironomids larvae (found in shallow water). Possibleexplanations include the ideas that fish species may bechoosing the same prey of different size (Di Prinzio et al.2013), feeding in different places or at different times(Holt 1987), or exhibit food partitioning or feedingmodes (Kusabs and Swales 1991). Despite the shorttime of co-existence between trout and galaxiids, some

galaxiids already have shown differences in timingof their feeding (Glova et al. 1992), changes indiet composition and specialization (Elgueta et al.2013), and restriction to shallow covered areas(Habit et al. 2010; Correa et al. 2012) concordantwith the possible avoidance of interaction withsalmonids.

Similarities in diet between trout and galaxiids evi-denced that both species might feed at the same trophiclevel (McHugh et al. 2012), but the segregation ofhabitat or prey resources contributes to long-term co-existence. In addition, because Puyen is a small fish, theprey it consumes are restricted to the mouth and stomachsize, which might generate a reciprocal trophic nichedisplacement, similar to that found between G. plateiand trout (Correa et al. 2012). Thus, despite the negativeeffect that trout can generate on these native fishes, thesmall size of Puyenmay be advantageous for differentialuse of resources and therefore facilitate the coexistencewith trout.

Because the Santa Cruz River is a homogeneousriver, with low primary and secondary production(Miserendino 2001) and hydrologically very stable(Tagliaferro et al. 2013), juvenile trout and Puyen mayselect different habitat at the mesohabitat level to avoidfood competition (see Penaluna et al. 2009). Althoughthe specific mechanisms are to be explored, ourstudy shows the potential for a diet overlap be-tween these two species, and possible changesalong this large river. Whereas there is evidencethat adult Rainbow Trout may feed on galaxiids,especially in lakes (Macchi et al. 1999; McDowall2003; Arismendi et al. 2009), we found no evi-dence of juvenile rainbow Rainbow Trout preda-tion on Puyen (see also Arismendi et al. 2012).This suggests that the trophic interaction betweenunderyearling Rainbow Trout and native galaxiidsshould be contemplated from a competition pointof view.

The contrasting distribution pattern of Puyen andRainbow Trout along the mainstem of the Santa CruzRiver may be related to the differential habitat require-ment of both species or displacement of galaxiids byRainbow Trout. Similar to New Zealand counterpart,Puyen in Patagonia evolved in isolation from salmonids,and even if it coexisted with other fish predators, (e.g.perch), its predator-avoidance behavior might not beefficient enough to withstand the predation by adultsalmonids. Recent studies explained that most of the

Table 3 SIMPER results between trout and puyen diets in up-stream and mid-stream areas. Average abundance of prey in diet oftrout and puyen, percentage of contribution to diet (Cont. %), andcumulative percentage (Cum. %). Preys that contributed less than2.5 % were excluded from the table

AverageAbundance

Upstream Trout Puyen Cont. % Cum. %

M. chiloeensis 2.32 0.16 18.0 18.0

Hyalella sp. 1.35 1.68 16.7 34.7

Simuliidae 0.84 0.74 11.5 46.3

L. jaffuelli 0.97 0.37 10.6 56.8

A. michaelseni 0.57 0.26 8.2 65.0

Andesiops sp. 0.22 0.16 5.4 70.4

K. kuscheli 0.43 0.11 5.3 75.7

Chironomidae 0.14 0.79 4.9 80.6

Hydrobiosidae 0.38 0.05 4.5 85.1

L. cekalovici 0.11 0.68 4.4 89.5

Adult Trichoptera 0.3 0,00 3.5 93.0

Mid-stream

Hyalella sp. 2.97 2.18 15.6 15.6

L. cekalovici 1.75 2.03 13.5 29.1

L. jaffuelli 1.92 0.7 13.0 42.1

M. chiloeensis 1.17 0.77 10.4 52.5

Simuliidae 0.99 0.57 7.1 59.6

Adult Trichoptera 0.18 0.57 5.7 65.3

A. michaelseni 0.41 0.21 5.1 70.4

K. kuscheli 0.31 0.21 4.9 75.3

Hydrobiosidae 0.24 0.12 3.5 78.8

Chironomidae 0.03 0.55 3.3 82.1

Andesiops sp. 1.01 0.06 3.0 85.1

Lymnaea sp. 0.46 0.08 3.0 88.1

Adult Ephemeroptera 0.06 0.48 2.3 90.4

Environ Biol Fish (2015) 98:1311–1325 1321

naturalized juvenile Rainbow Trout are descendent fromanadromous Rainbow Trout, which spawn at the mid-upstream area of the Santa Cruz River (Liberoff 2013),this might generate higher abundances of RainbowTrout and consequently a higher competition pressure.On the other hand, Puyen has both landlocked anddiadromous behavior (Carrea et al. 2013; Tagliaferroet al. 2014) that spawn downstream in estuarine zones

explaining the greater abundance of this species close tothe river mouth.

This study represents the first extensive researchconcerning diet composition of native galaxiids andRainbow Trout along the second largest river inPatagonia. Given the un-disturbed hydrological andproduction characteristics of this river, this researchprovides valuable baseline data on the feeding tactics

Fig. 5 Amundsen diagrams (see Amundsen et al. 1996) of feed-ing tactics for Puyen (G. maculatus) and Rainbow Trout(O. mykiss) at the three river areas. The upper panel representsthe explanatory diagram for interpretation of feeding tactics, niche

width contribution and prey importance. For the niche widthcontribution, HPC=high between-phenotype component to nichewidth, and WPC=high within-phenotype component to nichewidth

1322 Environ Biol Fish (2015) 98:1311–1325

of native and exotic fish to be used to contrast with otherlarge rivers affected by human activities like agriculture,urbanization or stream regulation.

Future considerations on habitat preference shouldalso be taken into account. Changes in microhabitatcould produce a higher interaction between trout andnative galaxiids (McIntosh et al. 1992); species habitatwidth might change in presence of the invader(Penaluna et al. 2009) and the wider the environmentalrange, the more likely the coexistence between specieswill be (Meszéna et al. 2006). Due to the low habitatvariability characteristic of the Santa Cruz River, whichis reflected in the structure of macroinvertebrate com-munities (Tagliaferro et al. 2013), we expect RainbowTrout and galaxiids to have reduced opportunities forhabitat segregation, and a larger potential for competi-tion and displacement in feeding habits.

One obstacle when studying remote places, likeSouthern Patagonia, is that previous research is scantand baseline data inexistent. Moreover, non-invaded,control areas are scarce. The existence of areas withdifferential relative abundances of Rainbow Trout pro-vided us with a means to evaluate the potential trophicinteraction and segregation between native and exoticfish in this large river and contributed to the understand-ing of the invasive fish role during the first year of life.

Acknowledgments Funded by Consejo Nacional deInvestigaciones Científicas y Tecnológicas and Agencia Nacionalpara la Promoción de la Ciencia y la Tecnología and ConsejoNacional de Investigaciones Científicas y Tecnológicas grant,

Argentina. M.T. was supported by CONICET Graduate Fellow-ship. Centro Nacional Patagónico (CENPAT-CONICET) providedsupport for the optic service. Ea. Río Bote, Ea. Condor Cliff, Ea.Rincón Grande, and Ea. Los Plateados provided logistic support.This research project was conducted under the animal care regu-lations of CONICET and was authorized by the Argentinean fishcare institution: “Subsecretaria de Pesca de Santa Cruz”.

References

Abrams P (1980) Some comments on measuring niche overlap.Ecology 61:44–49. doi:10.2307/1937153

Amundsen PA, Gabler HM, Staldvik FJ (1996) A newapproach to graphical analysis of feeding strategy fromstomach contents data – modification of the Costello(1990) method. J Fish Biol 48:607–614. doi:10.1006/jfbi.1996.0060

Arismendi I, Soto D, Penaluna B, Jara C, Leal C, León-Muñóz J(2009) Aquaculture, non-native salmonid invasions, and as-sociated declines of native fishes in lakes of northern ChileanPatagonia. Freshw Biol 54:1135–1147. doi:10.1111/j.1365-2427.2008.02157.x

Arismendi I, González J, Soto D, Penaluna B (2012) Piscivory anddiet overlap between two non-native fishes in southernChilean streams. Aust Ecol 37:346–354. doi:10.1111/j.1442-9993.2011.02282.x

Arismendi I, Penaluna B, Dunham JB, García C, de Leaniz D,Soto IF, Gomez-Uchida D, Gajardo G, Vargas PV, León-Muñoz J (2014) Differential invasion success of salmonidsin southern Chile: patterns and hypotheses. Rev Fish BiolFish 24:919–941. doi:10.1007/s11160-014-9351-0

Basulto S (2003) El largo viaje de los salmons: una crónicaolvidada, propagación y cultivo de especies acuáticas enChile. [The long salmon trip: a forgotten chronicle, propaga-tion, and cultivation of aquatic species in Chile]. MavalLimitada [Eds.], Santiago, Chile

Bøhn T, Amundsen PA, Sparrow A (2008) Competitive exclusionafter invasion? Biol Invasions 10:359–368. doi:10.1007/s10530-007-9135-8

Brunet F, Gaiero D, Probst JL, Depetris PJ, Gauthier-Lafaye F,Stille P (2005) δ13C tracing of dissolved inorganic carbonsources in Patagonian rivers (Argentina). Hydrol Process19(17):3321–3344. doi:10.1002/hyp.5973

Buria L, Albariño R, Díaz-Villanueva V, Modenutti B, Balseiro E(2007) Impact of exotic rainbow trout on benthic macroin-vertebrate community from Andean-Patagonian headwaterstreams. Arch Hydrobiol 168(2):145–154. doi:10.1127/1863-9135/2007/0168-0145

Carrea C, Cussac VE, Ruzzante DE (2013) Genetic and pheno-typic variation among Galaxias maculatus populations re-flects contrasting landscape effects between northern andsouthern Patagonia. Freshw Biol 58(1):36–49. doi:10.1111/fwb.12036

Chipps SR, Garvey JE (2007) Assessment of diets and feedingpatterns. In C.S. Guy and M.L. Brown, [Eds.]: Analysis andinterpretation of freshwater fisheries data. AmericanFisheries Society, Bethesda

Fig. 6 Ivlev Index of prey selectivity for Rainbow Trout andPuyen. Index value above 0.5 indicates possitive electivity, andbelow -0.5 indicates avoidance of prey

Environ Biol Fish (2015) 98:1311–1325 1323

Clarke KR (1993) Non-parametric multivariate analyses of chang-es in community structure. Aust J Ecol 18:117–143. doi:10.1111/j.1442-9993.1993.tb00438.x

Clarke KR, Gorley RN (2006) PRIMERV6: User Manual/Tutorial. PRIMER-E, Plymouth

Clarke KR,Warwick RM (2001) Change inMarine Communities:An Approach to Statistical Analysis and Interpretation, 2ndedn. PRIMER-E, Plymouth

Coghlan SM, Cainb GR, Ringler N (2007) Prey selection ofsubyearling Atlantic salmon and rainbow trout coexisting ina natural stream. J Fresh Ecol 22(4):591–608. doi:10.1080/02705060.2007.9664820

Colwell RK, Futuyma DJ (1971) On the measurement of nichebreadth and overlap. Ecology 52:567–576. doi:10.2307/1934144

Correa C, Bravo AP, Hendry AP (2012) Reciprocal trophic nicheshifts in native and invasive fish: salmonids and galaxiids inPatagonian lakes. Freshwater Biol 57(9):1769–1781. doi:10.1111/j.1365-2427.2012.02837.x

Crowl TA, Townsend CR, McIntosh A (1992) The impact ofexotic brown and rainbow trout on native fish: the case ofAustralasia. Rev Fish Biol Fish 2:217–241. doi:10.1007/BF00045038

Di Prinzio CY, Casaux RJ (2012) Dietary overlap among nativeand non-native fish in Patagonian low order streams. AnnLimnol-Int J Lim 48(1):21–30. doi:10.1051/limn/2011055

Di Prinzio CY, Miserendino ML, Casaux R (2013) Feeding strat-egy of the non-native rainbow trout, Oncorhynchus mykiss,in low-order Patagonian streams. Fish Manag Ecol 20:414–425. doi:10.1051/limn/2011055

Domínguez E, Fernández HR (eds) (2009) Macroinvertebradosbentónicos sudamericanos. Sistemática y biología.Fundación Miguel Lillo, Tucumán

Elgueta A, González J, Ruzzante DE, Walde SJ, Habit E (2013)Trophic interference by Salmo trutta onAplochiton zebra andAplochiton taeniatus in southern Patagonian lakes. J FishBiol 82(2):430–444. doi:10.1111/j.1095-8649.2012.03489.x

Flecker AS, Townsend CR (1994) Community-wide conse-quences of trout introductions in New Zealand streams.Ecol Appl 4:798–807. doi:10.2307/1942009

Giller PS (1984) Community structure and the niche. Chapmanand Hall, London

Glova GJ, Sagar P (1993) A further assessment of trophc andspatial inter-relations of galaxiids and salmonids in NewZealand. Ecol Freshw Fish 2:132–140. doi:10.1111/j.1600-0633.1993.tb00093.x

Glova GJ, Sagar PM, Näslund I (1992) Interaction for food andspace between populations of Galaxias vulgaris Stokell andjuvenile Salmo trutta L. in a New Zealand stream. J Fish Biol41:909–925. doi:10.1111/j.1095-8649.1992.tb02719.x

Greig H, McIntosh AR (2006) Indirect effects of predatory trouton organic matter processing in detritus-based stream foodwebs. Oikos 112(1):31–40. doi:10.1111/j.0030-1299.2006.14219.x

Habit E, Piedra P, Ruzzante DE, Walde SJ, Belk MC, Cussac VE,Gonzalez J, Colin N (2010) Changes in the distribution ofnative fishes in response to introduced species and otheranthropogenic effects. Glob Ecol Biogeogr 19:697–710.doi:10.1111/j.1466-8238.2010.00541.x

Hansson LA, Annadotter H, Bergman E, Hamrin SF, Jeppesen E,Kairesalo T, Luokkanen E, Nilsson PA, Sondergaard M,

Strand J (1998) Biomanipulation as an application of foodchain theory: constraints, synthesis, and recommendationsfor temperate lakes. Ecosystems 1:558–574. doi:10.1007/s100219900051

Holt RD (1987) On the relation between niche overlap and com-petition: the effect of incommensurable niche dimensions.Oikos 48:110–114. doi:10.2307/3565696

Hyslop EJ (1980) Stomach contents analysis—a review ofmethods and their application. J Fish Biol 17:411–429. doi:10.1111/j.1095-8649.1980.tb02775.x

Ivlev VS (1961) Experimental ecology of feeding of fishes. YaleUniversity Press, New Haven and London

Jones ML, Stockwell JD (1995) A rapid assessment procedure forthe enumeration of salmonine populations in streams. N AmJ Fish Manag 15:551–562. doi:10.1577/1548-8675.1995.0150551

Juncos R, Beauchamp DA, Vigliano PH (2013) Modeling preyconsumption by native and nonnative piscivorous fishes:implications for competition and impacts on shared prey inan ultraoligotrophic lake in Patagonia. Trans Am Fish Soc142:37–41. doi:10.1080/00028487.2013.730109

Kruskal JB (1964) Non-metric multidimensional scaling: a numer-ical method. Psychometrika 29:115–129. doi:10.1007/BF02289694

Kusabs IA, Swales S (1991) Diet and food resource partitioning inkoaro, Galaxias brevipinnis (Günther), and juvenile rainbowtrout, Oncorhynchus mykiss (Richardson), in two Taupostreams, New Zealand. N Z J Mar Fresh 25:317–325. doi:10.1080/00288330.1991.9516485

Lattuca ME, Battini MA, Macchi PJ (2008) Trophic interactionsamong native and introduced fishes in northern Patagonianoligotrophic lake. J Fish Biol 72:1306–1320. doi:10.1111/j.1095-8649.2008.01796.x

Lever C (1996) Naturalized fishes of the world. Academic, SanDiego

Liberoff AL (2013) Bases ambientales del comportamientomigratorio de la trucha arco iris (Oncorhynchus mykiss) delrío Santa Cruz. Tesis doctoral. Universidad Nacional deCórdoba

Lopretto EC, Tell G (1995) Ecosistemas de aguas continentales.Metodologías para su estudio. Ed. Sur

Macchi PJ, Cussac VE, AlonsoMF, Denegri MA (1999) Predationrelationships between introduced salmonids and the nativefish fauna in lakes and reservoirs in Northern Patagonia. EcolFreshw Fish 8:227–236. doi:10.1111/j.1600-0633.1999.tb00074.x

MacCrimmon HR (1971) World distribution of rainbow trout(Salmo gairdneri). J Fish Res Board Can 26:1699–1725.doi:10.1139/f71-098

Mack RN, Simberloff D, Lonsdale WM, Evans H, Clout M,Bazzaz FA (2000) Biotic invasions: causes, epidemiology,global consequences, and control. Ecol Appl 10:689–710

Marshall S, Elliott M (1997) A comparison of univariate andmultivariate numerical and graphical techniques for deter-mining inter- and intraspecific feeding relationships in estu-arine fish. J Fish Biol 51:526–545. doi:10.1111/j.1095-8649.1997.tb01510.x

McDowall RM (1990) New Zealand freshwater fishes, a naturalhistory and guide. Heinemann-Reed, Auckland

McDowall RM (1994) The origins of New Zealand’s Chinooksalmon, Oncorhynchus tshawytscha. Mar Fish Rev 56:1–7

1324 Environ Biol Fish (2015) 98:1311–1325

McDowall RM (2003) Impacts of introduced salmonids on nativegalaxiids in New Zealand upland streams: a new look at anold problem. Trans Am Fish Soc 132:229–238. doi:10.1577/1548

McHugh P, McIntosh A, Howard S, Budy P (2012) Niche flexi-bility and trout-galaxiid co-occurrence in a hydrologicallydiverse riverine landscape. Biol Invasion 14(11):2393–2406. doi:10.1007/s10530-012-0237-6

McIntosh AR (2000) Habitat-and size-related variations in exotictrout impacts on native galaxiid fishes in Zealand streams.Can J Fish Aquat Sci 57:2140–2151. doi:10.1139/cjfas-57-10-2140

McIntosh AR, Townsend CR, Crowl TA (1992) Competition forspace between introduced brown trout (Salmo trutta L.) andcommon river galaxiid (Galaxias vulgaris Stokell) in a NewZealand stream. J Fish Biol 41:63–81. doi:10.1111/j.1095-8649.1992.tb03170.x

McIntosh AR, McHugh PA, Dunn NR, Goodman JM, HowardSW, Jellyman PG, O’Brien LK, Nyström P, Woodford D(2010) The impact of trout on galaxiid fishes in NewZealand. N Z J Ecol 34:195–206

Meador MR, McIntyre JP, Pollock KH (2003) Assessing theefficacy of single-pass backpack electrofishing to character-ize fish community structure. Trans Am Fish Soc 132:39–46.doi:10.1577/1548-8659.2003.132.0039

Meszéna G, Gyllenberg M, Pásztor L, Metz JAJ (2006) Competitiveexclusion and limiting similarity: a unified theory. Theor PopulBiol 69:68–87. doi:10.1016/j.tpb.2005.07.001

Miserendino ML (2001) Macroinvertebrate assemblages inAndean Patagonian rivers and streams: environmental rela-tionships. Hydrobiologia 444:147–158

Moyle PB, Light T (1996) Biological invasions of fresh water:empirical rules and assembly theory. Biol Conserv 78:149–161. doi:10.1016/0006-3207(96)00024-9

Muotka T, Syrjänen J (2007) Changes in habitat structure, benthicinvertebrate diversity, trout populations and ecosystem pro-cesses in restored forest streams: a boreal perspective. FreshwBiol 52(4):724–737. doi:10.1111/j.1365-2427.2007.01727.x

Pascual MA, Ciancio JE (2007) Introduced anadromous salmo-nids in Patagonia: risks, uses, and a conservation paradox. InT. M. . Bert (Ed.), Ecological and Genetic ImplicationsofAquaculture Activities (pp. 1–19). Klewer Publications

Pascual M,Macchi P, Urbanski J, Marcos F, Riva-Rossi C, NovaraM, Dell’Arciprete P (2002) Evaluating potential effects ofexotic freshwater fish from incomplete species presence-absence data. Biol Invasions 4:101–113. doi:10.1023/A:1020513525528

Penaluna BE, Arismendi I, Soto D (2009) Evidence of interactivesegregation between introduced trout and native fishes innorthern Patagonian rivers. Chil Trans Am Fish Soc 138:839–845. doi:10.1577/T08-134.1

Schöener TW (1970) Non synchronous spatial overlap of lizardsin patchy habitats. Ecology 51:408–418. doi:10.2307/1935376

SimonKS, Townsend CR (2003) Impacts of freshwater invaders atdifferent levels of ecological organisation, with emphasis onsalmonids and ecosystem consequences. Freshw Biol 48(6):982–994. doi:10.1046/j.1365-2427.2003.01069.x

Soto D, Arismendi I, Di Prinzio C, Jara F (2007) Establishment ofChinook salmon (Oncorhynchus tshawytscha) in Pacific ba-sins of Southern South America and its potential ecosystemimplications. Rev Chil Hist Nat 80(1):81–98. doi:10.4067/S0716-078X2007000100007

Tagliaferro MB (2014) Estructura espacial, temporal y trófica delas comunidades acuáticas del río Santa Cruz. (Spatial, tem-poral, trophic structure of the acquatic communitiesof theSanta Cruz River). PhD Thesis - Universidad de BuenosAires-Facultad de Ciencias Exactas y Naturales. 195pp

Tagliaferro MB, Miserendino ML, Liberoff AL, Quiroga P,Pascual MA (2013) Dams in the last large free-flowing riversof Patagonia, the Santa Cruz River, environmental features,and macroinvertebrate community. Limnologica 43:500–509. doi:10.1016/j.limno.2013.04.002

Tagliaferro M, Quiroga A, Pascual M (2014) Spatial pattern andhabitat requirements of galaxias maculatus in the last Un-interrupted large river of Patagonia: a baseline for manage-ment. Environ Nat Resour Res 4(1):54–64. doi:10.5539/enrr.v4n1p54

Tilman D (1987) The Importance of the Mechanisms ofInterspecific Competition. Am Nat 129(5):769–774. doi:10.1086/284672

Townsend CR (2002) Individual, population, community, andecosystem consequences of a fish invader in New Zealandstreams. Conserv Biol 17(1):38–47. doi:10.1046/j.1523-1739.2003.02017.x

Tulian EA (1908) Acclimatization of American fishes inArgentina. Bull. U S Bur Fish 18(2):957–965. doi:10.5962/bhl.title.40557

Vargas PV, Arismendi I, Lara G, Peredo S, Millar J (2010)Evidence of niche overlap between juvenile introducedChinook salmon (Oncorhynchus tshawytscha, Walbaum,1792) and native catfish (Trichomycterus areolatus,Valenciennes, 1842) in the Allipén river, Chile. Rev BiolMar Oceanogr 45(2):285–292

Wallace RK Jr (1981) An assessment of diet-overlap indexes.Trans Am Fish Soc 110:72–76. doi:10.1577/1548-8659(1981)110%3C72:AAODI%3E2.0.CO;2

Warwick RM, Clarke KR (1993) Increased variability as a symp-tom of stress in marine communities. J Exp Mar Biol Ecol172:215–226. doi:10.1016/0022-0981(93)90098-9

Welcomme RL (1988) International introductions of inland aquat-ic species. Food and Agriculture Organization of the UnitedNations, Fisheries Technical Paper, 294: 1–318. Rome, Italy

Young KA, Dunham JB, Stephenson JF, Terreau A, Thailly AF,Gajardo G, García de leaniz C (2010) A trial of two trouts:comparing the impacts of rainbow and brown trout on anative galaxiid. Anim Conserv 13:399–410. doi:10.1111/j.1469-1795.2010.00354.x

Environ Biol Fish (2015) 98:1311–1325 1325